Largazole, a cyclic depsipeptide originally isolated from a marine cyanobacterium Symploca sp., has been shown to be an anti-tumor agent (Taori et al. 2008). Largazole specifically targets histone deacetylases whose dysfunction is often associated with a variety of human tumors. Largazole has been shown to: (i) display nM GI50 values against a variety of cell lines (eg MDA-MB-231 mammary carcinoma cells, GI50=7.7 nM; U2OS fibroblastic osteosarcoma cells, GI50=55 nM; HT29 colon cells, GI50=12 nM; IMR-32 neuroblastoma cells (Taori et al. 2008), GI50=16 nM) (Taori et al. 2008), (ii) display differential activity between transformed and non-transformed cells (Nasveschuk et al. 2008; Taori et al. 2008; Ungermannova 2010) and (iii) is structurally simpler and possibly more tractable synthetically than the other depsipeptides.
Although, the largazole molecule is a proven antitumor agent, there is always a need for improved structural analogs that lead to improved HDAC inhibition properties, toxicity and physiochemical profiles resulting is improved cancer therapies.
It has been known for years that DMSO and butyrate, two known relatively nonspecific inhibitors of HDACs, can induce certain leukemia cells to differentiate and suppress neoplastic growth (Sato et al. 1971; Leder et al. 1975).
In recent years, HDACs and histone acetylases (HATs) have become widely recognized as key players in regulating transcription (Minucci and Pelicci 2006). Acetylation of lysines in the histone H3 and histone H4 tails is strongly correlated to chromatin states that are ready for transcription, or that are part of actively transcribed genomic regions (Allfrey et al. 1964). Acetylation of histones has also been correlated with other important cellular functions including chromatin assembly, DNA repair, and recombination.
There are 18 HDAC enzymes in the human genome that can be classified into four classes (Lane and Chabner 2009). Classes I, II and IV all contain a zinc (Zn2+) molecule in their active site (Table 1 adapted from (Lane and Chabner 2009)).
Because of their important role in regulating transcription and disruptions of their regulation in tumor cells, it has been postulated that inhibition o HDAC could be an effective way for cancer therapeutics. Consequently, there has been substantial development in inhibitors of HDAC enzymes (HDACi) as potential anti-cancer drugs (Marks). The clinical relevance of this attention to HDACi is warranted and has recently been underscored by the introduction of vorinostat (Zolinza™, Merck, also widely known as SAHA=suberoylanilide hydroxamic acid) for the treatment of cutaneous T-cell lymphoma in late 2006 and more recently Romidepsin (FK228) (Marks).
The catalytic activity of HDAC contains features from both serine protease and metalloprotease enzymes. On the basis of the crystal structures of HDAC8 and a bacterial histone deacetylase-like protein (HDLP), the mechanism for the deacetylation reaction has been proposed (Finnin et al. 1999; Somoza et al. 2004; Vannini et al. 2004). There is a deep, tube-like narrow pocket that expands at the bottom and an internal cavity that borders the pocket (FIG. 3a of Finnin). The inside of the tube is comprised of hydrophobic and aromatic residues. The zinc ion is situated at the bottom of the pocket and Zn2+ and His 142, acting as a general base, activate the water molecule for nucleophilic attack on the carbonyl group of the substrate. This would result in a tetrahedral carbon that is stabilized by the formation of a hydrogen bond with Tyr 306 and a general acid His 143 that protonates the lysine leaving group, yielding the acetate and lysine products. Both His 142 and His 143 fit in the Asp 166-His 131 charge-relay system, which is proposed to modulate the basicity of the His residues (FIG. 5 of Finnin) (Finnin et al. 1999).
The mode of action of a majority of HDAC inhibitors is to mimic the substrate interactions with the deacetylase, thus preventing the entry of the acetylated lysine residue located on the tail of the histone protein. All small molecule histone deacetylase inhibitors share three structural elements that contribute to HDAC inhibition: (1) a surface recognition domain which is anchored at the rim of the HDAC's tube-like pocket, (2) a zinc binding site, (3) a linker region that connects the surface recognition domain to the zinc binding site (Finnin et al., 1999). FIG. 1.5 (adapted form (Newkirk et al. 2009)) shows a general pharmacophore model of several known HDAC inhibitors.
While SAHA exerts its anti-cancer activity at least in part by the modulation of HDACs in a direct fashion by coordination of the Zn2+ ion in the active site of the enzyme by the terminal hydroxamic acid, it displays poor selectivity among the 3 classes of HDACs in part due to its structure simplicity (Minucci and Pelicci 2006; Lane and Chabner 2009). It has been generally accepted that Class I HDACs are more relevant to cancer therapy and poor selectivity of HDAC inhibitors are responsible for chronic toxicities (Minucci and Pelicci 2006; Lane and Chabner 2009). In search for more specific Class I HDAC inhibitors has led to the discovery of a number of natural product depsipeptides including FR901375 (Koho 1991), FK228 (Ueda et al. 1994a), spiruchostatin A (Masuoka et al. 2001), and the very recently isolated largazole (Taori et al. 2008) (FIG. 2). These natural products known as Depsipeptides share a common feature in that they all contain an (3S,4E)-3-hydroxy-7-mercapto-4-heptenoic acid side chain (Newkirk et al. 2009). A free sulfhydryl (thiol) needs to be exposed in this class of compounds to unleash the inhibitory activity of HDAC as the thiol coordinates the active site Zn2+ ion to prevent catalysis.
The Zn2+ binding moiety of largazole is inactive unless the thioester is removed by hydrolysis. It has been demonstrated that largazole-thiol is the active specie that potently inhibit HDACs (Bowers et al. 2008; Ying et al. 2008b). Thus largazole is most likely a pro-drug that becomes activated by esterase/lipases upon uptake into cells or conjugated to a carrier/transport protein and reduced to thiol intracellularly. There have been significant developments in prodrug design to improve physicochemical and pharmacokinetic properties of biologically potent lead compounds (Rautio et al. 2008). However, these strategies have not been systematically applied to largazole.